FDI has Promoted Diversification and Sophistication of Goods and Services Exports

0%

2%

4%

6%

8%

10%2

00

0

20

01

20

02

20

03

20

04

20

05

20

06

20

07

20

08

20

09

20

10

20

11

20

12

20

13

20

14

20

15

Shar

e

Goods exports H-H concentration index*FDI (Share of GDP)Trend of FDI (Share of GDP)

0%

10%

20%

30%

40%

50%

60%

70%

20

00

20

01

20

02

20

03

20

04

20

05

20

06

20

07

20

08

20

09

20

10

20

11

20

12

20

13

20

14

20

15

Electric&Electronic+MedicalDevices/All Manufacturing

Information&Computing+BusinessServices/All Services

Costa Rica: FDI Inflows and Goods Exports Diversification Costa Rica: Services and Goods Exports Sophistication

Sources: Authors, using data from BCCR and PROCOMER. Note:* 6-digit HS aggregation

Costa Rica’s TFP has been Stagnating

Source: Central Bank of Costa Rica. Data from PWT Version 9.0

Costa Rica: Total Factor Productivity relative to USA

Transportation, IT and

communications

Business services Electricity, gas and water Public administration

services

Construction Financial services and

insurance

Trade, restaurants and hotels Communal, social and

personal services

Costa Rica: Labour Productivity by Sector, 2001-15 (2001 = 100)

Source: Mulder et. al. 2016. Análisis del Crecimiento de la Productividad en Costa Rica, 2001-2015

Heterogeneous Labour Productivity across Services Activities

9.1

10.9

8.1

2.3

4.4

1.0

3.0 2.8 3.1

1.8 1.2

2.2

1991-2015 1991-2000 2000-2015

FTZ manuf. Out-of-FTZ manuf.

All manufacturing Overall economy

Costa Rica: Real Annual Growth of Value Added per Worker (1991-2015)

Dual Performance: Labour Productivity Growth Differs Significantly between FTZ Firms and Local Firms

Source: Medaglia, 2015. Balance de Zonas Francas: beneficio neto del régimen para Costa Rica, 2011-2015

Research Approach

• Question: Is there evidence of FDI spilling over domestic firms’ productivity?

• Hypothesis: Backward linkages and foreign presence positively affects domestic firms’ productivity

• Methodology: Fixed effects (FE) and random effects (RE) panel data equation

• Data • Firm-level data for period 2008-15, formal sector • Source: Central Bank of Costa Rica’s data set constructed using

administrative records from several official primary sources • Variables: revenues, total net assets, employment, imports, exports,

foreign capital share, sectoral classification (4-digit ISIC v.4), transactions between firms, location

• Building also on results by Alfaro, Manelici and Vásquez (2017) for data on transactions between FDI and local firms and for tracking FDI firms in CR.

Ruan & Ugur (2005)’s model for manufacturing

𝑌

𝐿 𝑖𝑗𝑡= 𝛽0 + 𝛽1𝐹𝑃𝑗𝑡 + 𝛽2

𝐾

𝐿 𝑖𝑗𝑡+ 𝛽3

𝐿𝑠

𝐿𝑢 𝑖𝑗𝑡

+ 𝛼𝑗 + 𝛼𝑡 + 𝑒𝑖𝑗𝑡

Where,

•𝑌

𝐿 𝑖𝑗𝑡: labour productivity of firm i in sector j in year t. Ratio of revenues (Y) to total employment (L)

• 𝐹𝑃𝑗𝑡: foreign presence. Share of employment accounted by all foreign-owned plants in the relevant sector

•𝐾

𝐿 𝑖𝑗𝑡: capital intensity. Ratio of total net assets by total employment in the firms

•𝐿𝑠

𝐿𝑢 𝑖𝑗𝑡: labour quality. Ratio by skilled workers (Ls) to unskilled workers (Lu)

• Sector, province and time dummies are included

Panel Data Equation: Specification No.1

Panel Data Equation: Specification No.2

Haller (2014)’s model for services

𝑌

𝐿 𝑖𝑗𝑡= 𝛽0 + 𝛽1𝐹𝑃𝑗𝑡−1 + 𝛽2𝑒𝑥𝑝𝑖𝑡−1 + 𝛽3 𝑒𝑥𝑝𝑖𝑡−1 ∗ 𝐹𝑃𝑗𝑡−1 +

𝛽4𝑖𝑚𝑝𝑐𝑜𝑚𝑗𝑡−1 + 𝛽5𝐻𝐻𝐼𝑗𝑡 + 𝛽6∆𝑖𝑔𝑟𝑗𝑡−1 + 𝛼𝑗 + 𝛼𝑡 + 𝑒𝑖𝑡

Where,

•𝑌

𝐿 𝑖𝑗𝑡: labour productivity of firm i in sector j in year t

• 𝐹𝑃𝑗𝑡−1: foreign presence. Lagged share of employment accounted by all foreign-owned plants in the relevant sector j

• 𝑒𝑥𝑝𝑖𝑗𝑡−1: lagged export status of firm i. Dummy variable assumes value 1 if firm registered exports that year

• 𝑖𝑚𝑝𝑐𝑜𝑚𝑗𝑡−1: import competition from abroad. Lagged imports share in domestic consumption by 2-digit ISIC code

• 𝐻𝐻𝐼𝑗𝑡: Herfindahl-Hirschman index to control for product market concentration in sector j (3-digit ISIC code)

• ∆𝑖𝑔𝑟𝑗𝑡−1: lagged revenue growth in sector j, to control for firms’ willingness to join fast growing industries

• Sector, province and time dummies are included

Foreign Presence in Manufacturing and Services

0%

10%

20%

30%

40%

50%

60%

2008 2009 2010 2011 2012 2013 2014 2015

Shar

e

Employees Revenues Firms

0%

10%

20%

30%

40%

50%

60%

2008 2009 2010 2011 2012 2013 2014 2015

Employees Revenues Firms

Sources: Authors, using data from BCCR

Foreign Firms Share in Manufacturing, 2008-15 Foreign Firms Share in Services, 2008-15

Firm and Sector-level Descriptive Statistics

Firm-level variables Foreign firms Local firms

All Manufacturing Services

Variable Mean Stand.

Deviation Mean

Stand. Deviation

Mean Stand.

Deviation Mean

Stand. Deviation

ln(Y/L) 17.71 1.20 17.15 1.05 17.05 0.90 17.17 1.07

ln(K/L) 17.36 1.59 16.39 1.60 16.15 1.37 16.42 1.63

Ls/Lu 40.82 121.81 3.14 15.95 2.19 12.11 3.27 16.38 Share of sales to FO firms 0.15 0.23 0.10 0.22 0.13 0.23 0.09 0.21 Exporter 0.46 0.50 0.05 0.22 0.15 0.36 0.04 0.20 No. of observations 6,025a 180,811* 20,920* 159,891*

Variable 2-digit aggreg.

Foreign presence

Import competition

3-digit aggreg.

HHI Revenues

growth rate

Year

No. of sectors

Mean Stand. Dev.

Mean Stand. Dev.

No. of sectors

Mean Stand. Dev.

Mean Stand. Dev.

Manufacturing 2008 19 0.48 0.28 0.43 0.22 35 0.33 0.24 0.004 0.02

2015 20 0.45 0.30 0.34 0.17 37 0.28 0.24 0.0004 0.01

Services 2008 41 0.23 0.21 0.12 0.12 81 0.17 0.16 0.003 0.016

2015 41 0.25 0.12 0.12 0.12 87 0.13 0.16 0.004 0.018

Sector-level variables

Source: Authors, using data from BCCR. Notes: a) Minimum number of observations available.*Regression sample.

Results: Specification No.1 (Ruan & Ugur) Applied to Domestic Manufacturing Firms

Productivity Spillovers in Manufacturing at 2-Digit Level

Model (1) (2) (3) (4) (5) (6) FE RE FE RE FE RE

Foreign presence 0.070 0.065 0.070 0.065 [0.132] [0.131] [0.132] [0.131]

Share of sales to FO firms 0.016 0.056 0.016 0.056 [0.045] [0.039] [0.045] [0.039]

ln(K/L) 0.215*** 0.238*** 0.215*** 0.238*** 0.215*** 0.238*** [0.013] [0.011] [0.013] [0.011] [0.013] [0.011]

Ls/Lu 0.001*** 0.001*** 0.001*** 0.001*** 0.001*** 0.001*** [0.000] [0.000] [0.000] [0.000] [0.000] [0.000] Year dummies Yes Yes Yes Yes Yes Yes Sector dummies - Yes - Yes - Yes Province dummies - Yes - Yes - Yes No. of observations 20,920 20,920 20,920 20,920 20,920 20,920 Number of firms 4,059 4,059 4,059 4,059 4,059 4,059 Prob > F 0,0 0,0 0,0 Prob > chi2 0,0 0,0 0,0

Source: Authors, using data from BCCR. Notes: Standard errors in brackets. *** p<0.01, ** p<0.05, * p<0.1

Results: Specification No.1 (Ruan & Ugur) Applied to Domestic Services Firms

Model (1) (2) (3) (4) (5) (6) FE RE FE RE FE RE

Foreign presence 0.430*** 0.488*** 0.433*** 0.488*** [0.099] [0.097] [0.099] [0.097]

Share of sales to FO firms -0.079*** 0.004 -0.079*** 0.004 [0.021] [0.017] [0.021] [0.017]

ln(K/L) 0.209*** 0.226*** 0.209*** 0.226*** 0.208*** 0.226*** [0.005] [0.004] [0.005] [0.004] [0.005] [0.004]

Ls/Lu 0.000 0.000*** 0.000 0.000*** 0.000 0.000*** [0.000] [0.000] [0.000] [0.000] [0.000] [0.000] Year dummies Yes Yes Yes Yes Yes Yes Sector dummies - Yes - Yes - Yes Province dummies - Yes - Yes - Yes No. of observations 159,891 159,891 159,891 159,891 159,891 159,891 Number of firms 37,523 37,523 37,523 37,523 37,523 37,523

Prob > F 0,0 0,0 0,0 Prob > chi2 0,0 0,0 0,0

Productivity Spillovers in Services at 2-Digit Level

Source: Authors, using data from BCCR. Notes: Standard errors in brackets. *** p<0.01, ** p<0.05, * p<0.1

Results: Specification No.2 (Haller) Applied to Domestic Manufacturing Firms

Model (1) (2) (3) (4)

FE RE FE RE

Foreign presence (lag): Non-Exp -0.017 -0.035

[0.115] [0.115]

Foreign presence (lag): Exp -0.004 -0.027

[-0.033] [-0.213]

Share of sales to FO firms: Non-Exp 0.024 0.081*

[0.051] [0.045]

Share of sales to FO firms: Exp -0.065 -0.046

[0.070] [0.066]

Impcomp (lag) 0.148 0.146 0.144 0.122

[0.105] [0.105] [0.105] [0.104]

HHI -0.410*** -0.417*** -0.416*** -0.423***

[0.135] [0.138] [0.129] [0.132]

Revenues growth rate (lag) 0.072 0.054 -0.269 -0.288*

[0.322] [0.314] [0.177] [0.172]

Year dummies Yes Yes Yes Yes

Sector dummies - Yes - Yes

Province dummies - Yes - Yes

No. of observations 19,242 19,242 19,343 19,343

Number of firms 3,728 3,728 3,728 3,728

Prob > F 0,0 0,0

Prob > chi2 0,0 0,0

Productivity Spillovers in Manufacturing at 3-Digit Level

Source: Authors, using data from BCCR. Notes: Standard errors in brackets. *** p<0.01, ** p<0.05, * p<0.1

Results: Specification No.2 (Haller) Applied to Domestic Services Firms

Productivity Spillovers in Services at 3-Digit Level

Model (1) (2) (3) (4) FE RE FE RE

Foreign presence (lag): Non-Exp 0.133* 0.169** [0.074] [0.073]

Foreign presence (lag): Exp -0.002 0.011

-0.0160 0.0958

Share of sales to FO firms: Non-Exp -0.096*** -0.021

[0.023] [0.019]

Share of sales to FO firms: Exp -0.170 -0.131 [0.057] [0.055]

Impcomp (lag) -0.190** -0.197** -0.188** -0.195** [0.093] [0.091] [0.091] [0.089]

HHI -0.158** -0.176*** -0.122* -0.136** [0.068] [0.067] [0.066] [0.066]

Revenues growth rate (lag) -0.220 -0.226 0.081 0.078 [0.158] [0.155] [0.119] [0.117]

Year dummies Yes Yes Yes Yes

Sector dummies - Yes - Yes

Province dummies - Yes - Yes

No. of observations 139,703 139,703 140,601 140,601

Number of firms 32,367 32,367 32,452 32,452

Prob > F 0,0 0,0

Prob > chi2 0,0 0,0

Source: Authors, using data from BCCR. Notes: Standard errors in brackets. *** p<0.01, ** p<0.05, * p<0.1

Results: Mixed Specification combining Ruan & Ugur + Haller, Applied to Domestic Manufacturing and Services Firms

Model

Manufacturing Services

(1) (2) (3) (4)

FE RE FE RE

Foreign presence (Non- exp) 0.102 0.056 0.428*** 0.480***

[0.137] [0.135] [0.098] [0.096]

Foreign presence (Exp) 0.171 0.0861 0.563*** 0.645***

[0.168] [0.161] [0.171] [0.158]

Share of sales to FO firms 0.008 0.043 -0.091*** -0.004

[0.044] [0.039] [0.022] [0.018]

Impcomp (lag) 0.152 0.089 -0.172** -0.190**

[0.095] [0.093] [0.084] [0.082]

IHH -0.310** 0.042 -0.060 0.031

[0.124] [0.079] [0.062] [0.056]

Revenues growth rate (lag) 0.094 0.056 -0.143 -0.153

[0.283] [0.277] [0.131] [0.130]

ln(K/L) 0.213*** 0.234*** 0.203*** 0.223***

[0.014] [0.012] [0.005] [0.004]

Ls/Lu 0.001*** 0.001*** 0.000** 0.000***

[0.000] [0.000] [0.000] [0.000]

Year dummies Yes Yes Yes Yes

Sector dummies - Yes - Yes

Province dummies - Yes - Yes

No. of observations 19,344 19,344 140,601 140,601

Number of firms 3,736 3,736 32,452 32,452

Source: Authors, using data from BCCR. Notes: Standard errors in brackets. *** p<0.01, ** p<0.05, * p<0.1

Productivity Spillovers in Manufacturing and Services at 2-Digit Level

Comparison of Results

Specification Variable

Manufacturing Services

(1) (2) (3) (2)

FE RE FE RE

Ruan&Ugur

Foreign presence 0.070 0.065 0.433*** 0.488***

[0.132] [0.131] [0.099] [0.097]

Share of sales to FO firms 0.016 0.056 -0.079*** 0.004

[0.045] [0.039] [0.021] [0.017]

Haller

Foreign presence (lag) Non-Exp -0.017 -0.035 0.133* 0.169**

[0.115] [0.115] [0.074] [0.073]

Foreign presence (lag) Exp -0.004 -0.027 -0.002 0.011

[-0.033] [-0.213] -0.0160 0.0958

Combination

Foreign presence (Non- exp) 0.102 0.056 0.428*** 0.480***

[0.137] [0.135] [0.098] [0.096]

Foreign presence (Exp) 0.171 0.0861 0.563*** 0.645***

[0.168] [0.161] [0.171] [0.158]

Share of sales to FO firms 0.008 0.043 -0.091*** -0.004 [0.044] [0.039] [0.022] [0.018]

Source: Authors, using data from BCCR. Notes: Standard errors in brackets. *** p<0.01, ** p<0.05, * p<0.1

Findings of Productivity Spillovers in Manufacturing and Services througth the Specifications

Ongoing Work

• Run estimations at sub-sector level (i.e. plastics, metal mechanic works, business services, IT services, etc.)

• Identify comparable samples of domestic firms to control for differences when estimating the effect from linkages with FDI

Conclusions and Policy Implications

Conclusions • So far, no evidence of FDI spillovers on domestic manufacturing firms’ productivity

at the aggregate level • So far, evidence of FDI spillovers on domestic services firms’ productivity at the

aggregate level • More research is needed with a higher level of disaggregation, for example by

applying the model to specific sub-sectors, particularly in services

Policy Implications • Need to deploy a public policy agenda for domestic firms’ capacity building

oriented to boost their absorptive capacity and linking with FDI